Rick Altman

Toward a Theory of the History

of Representational Technologies

Recent interest in the technology of representational media, coupled

with a growing concern to theorize cinema history, has made it possi-
ble, finally, to contemplate a history of representational technolo-
gies/ l. While recent criticism has succeeded in opening this new
territory, however, it has simultaneously staked out claims and esta-
blished practices "".hichthreaten to close off this fertile area before it
can be permanently and intelligently settled. The problems which these
recent efforts have encountered are instructive in themselves, however.
Careful consideration of three such problems will lead me here to a new
hypothesis regarding the history of representational technologies.
1. Technique/Technology
Jean-Louis Comolli can hardly be made to carry the full responsabi-
lity for the general tendency of cinema theorists to conflate the concepts
of technique and technology (though his« Technique et ldeologie »is to
be sure complicitous in this affair). The three languages which contri-
bute most regularly to the realm of cinema theory ,-English, French,
and German- are already marked by a certain confusion in those
terms derived from the Greek techne (skill, art, or craft). In English, the
adjective «technical » refers alternately to technique and to technology,
while the noun, «technician» has come to replace the more logical
«technologist» to designate someone who works with technology.

/ I. Jean-Louis Comolli's landmark 1971-72 essay «Technique et ldeologie»

( Cahiersdu Cinema 229, 4-21 ; 230, 51-57; 231, 42-49; 233, 39-45; 234-S, 94-100; 241,
2().24) is followed by Ron Burnett,• Film/Technology/Ideology». Cine-tracts I, 1977,
6,.14; Edward Branigan, «Color and Cinema: Problems in the Writing of History »1 in
Allan Almendarez, Lafferty, eds., Film Reader 4, Evanston, Northwestern univ.,
1979 • 16,.34; Charles H. Harpole, «Ideological and Technological Determinism in
Deep-Space Cinema Images: Issuesin Ideology, Technological His!O(Y,and Aesthe-
tics», Film Quater/y 33, no. 3, Spring 1980, 11-22; Tere~a,de Lauretls and Sl<:P.hen
Heath. eds. TheCinematic Apparatus, New York, S' Martins Press~1980.Inadd1tmn,
the work of Patrick Ogle. Douglas Gomery. and William Lafferty deserves to be cited.
112 Nick A flman

French tends to use the single term «technique» to designate the entire
range of meanings. German is still more perverse, often using «Techno-
logie » to mean technique and « Technik » to mean technology /2. This
potential confusion has been further compounded by the choice of
deep-focus photography as the major proving ground of theories regar-
d ind the relationship between technology and history, for the produc-
tion of a deep-focus image -like many other filmic phenomena-
depends on a combination of technical and technological concerns.
While technological changes in the late thirties (availability of new
lenses) make deep-focus photography easier and more economical to
achieve, anyone who has ever used a camera knows that a depth-of-
focus choice must be made every time the shutter is tripped. The
difference between an exposure made at f 5.6with a speed of l /250and
another made at f 16with a speed of l/ 30 isa question of technique, not
of technology ; the latter image may be a deep-focus image, the former
cannot possible be. Indeed, given sufficient light, a cameraman and
director may choose to treat any shot with great depth of focus, with or
without the technological changes commonly associated with deep
focus. It is thus hardly surprising to find that articles which touch on
the topic of deep-focus photography characteristically alternate bet-
ween technical and technological concerns without making any distinc-
tion whatsoever between the two.
So what? one might well ask. Why so much commotion over a
simple question of vocabulary? Precisely because, I would claim, this is
no simple question of vocabulary, but a fundamental problem in the
theory of history. As we clearly see from the example of deep-focus
photography, the basic configuration is this: when the same result is
produced by two recognizably different causes, then critics feel justified
m conflating the terms habitually employed to distinguish one of those
causes from the other. When we scratch this simple surface we find a
logical application of one of the ground rules of semiotic analysis,
namely commutation: if substitution of one sound unit (or constituent
sense unit) for another makes no difference to the meaning (or higher-
level sense unit), then we say that the language (or text) recognizes no
difference in the alternate units. Following this reasoning, a generation
of film historians has taken technique and technology to be interchan-
geable notions because they yield similar results. This easy transferral
of methodology from synchronic semiotic analysis to the discourse of
history simply will not do. We must learn to use the familiar commuta-
tion test in a different way when we come to practice history. Whereas

/2. This confusion is especially evident in the work of Adorno. See Miriam Hansen,
« Introduction to Adorno, 'Transparencies on Film' (1966) >>,New German Critique
24-25. Fall-Winter 1981-82, 186-205.
Toward n theory of the hislory of representational technologies I 13

synchronic analysis quite properly takes similarity of effect or sense as

an indicator of functional identity, history refuses the hypothesis of
synchrony and thus must begin its analysis by recognizing that diffe-
rent causes may produce the same effect. Instead of erasing difference
in the name of functional identity, history must ask a different series of
questions. Under what circumstances does one cause produce the effect
in question? Under what circumstances the other? What relationship
obtains between the two?
In short, to conflate the notions of technique and technology is to
destroy the possibility of understanding technical/technological his-
tory in a fully dialectical manner. If the two domains are taken to be
coterminous and ultimately identical, then their effect on each other
will necessarily remain invisible to our analysis. This would be espe·
cially regrettable given the importance of the technique/technology
dialectic throughout the history of representation. It is certainly no
secret, for example, that one of the prime 1t1oversof technological
development lies in the economic interest of automatizing (i.e. reducing
the production time of) those techniques which have become consecra-
ted by tradition. Even before the industrial revolution and its camera,
the Renaissance had produced numerous contraptions assuring « An
Easy Method of Representing Natural Objects According to the Rules
of Perspective», as the French Jesuit Jean Dubreuil would have it/ 3. In
fact, throughout the history of the camera, this automatization of
accepted technique by new technology continues the operate. During
the 1840s, standard procedure called for the photographer to immobi-
lize his client with a head brace. By the end of the century, rapid film
stock had transferred the process of immobilizing the subject from an
accepted photographer's technique to an automatic part of photogra-
phic technology. A similar analysis would show how the development
of directional ·microphones in the 1930s consecrated -and
automatized- persistent attempts throughout the early years of sound
cinema to limit and concentrate the range of existing mikes/4.

/ 3. lt is from this work that the frontispiece of Como!H·s first installment is drawn. In
passing it is perhaps worth noting that ComoUi truncates the plate, removing the
separate dose-ups of the devlces which make it possible for the depicted painter to
produce a perspective drawing automatically. In other words, Comolli's version seems
to rcfor to the technique of perspective alone, while the complete drawing clearly refers
to the te{;hnologizing of that technique as well. {See page 110 for a reproduction of the
comp1ete engraving).
J4. It is interesting to note similarity between Gomery's proposed stages of technolo-
gical deployment and the common pattern outlined he.re. Whereas Gomerts first stage
(invention) is largely technological, however. mine recognizes the possibility of inven-
tion thrnugh technique. In dealing with the technologization of technique~ however, I
do recognize the same economic impulse that characterizes the move from invention to
innovation for Gomcry,
ll4 Rick Airman

...~-

l
•

"h Very cud(HH Md:Md ct JawLt>g ~II P,ripc,:fr,t$ if>lb¢. m,m .1u1uri! maflntt, without®·
$t'tYin.gtht' Rut.,:· F,om Tin Pr,:c/;u i,j Ptnpu,;,;.,; w, An E11s1,~fr,had <>/~~f>U>lnting
N"umd 01,J(tlf Auorditr& /4 iht P;:,l~J 'If Pff!}'ntiu, wdttten in fu:rich by;\ J¢tuit of hth
(L<»i<Mn:Th<Je&yt.,, and j<!hn 1.myklc,M OCC X.XVI).
J.,,,.Duhrcuil, ih<:unn~mN lu!Mr, txpb!m rhu in the c.~guvi11g (A) i< a rinc nf g!a,, 1h;1l
~lip-; ir,w 1he ft.l""! {BC). !l:) ;, tn tdjustibtc 1ighl: vt<K, wiih a minute f('<i'{> hclc lt ii< !-Op,
thi1 litt inW 1h<,N.,c BO af l:is in;trum,:,-ir. Th{ 1rt!11 {F} u~>:d the oudin.:~ ,of wl:m he ~i
1.1pon ,•in<k;·,, (G). "EYte,jb<ldykll<Y"·a off, what it thu~ -0n th<:c1~,s," the
how to iakt. or ~<.>py
Jcwir t4nduJu, ··''fi1 bt~r to duw dit lbtt ;u,d figut¢S on the Ghtt with Pm ;;nJ fr,k: th~n
"~lrir.g th!! Back,1i<:kof th1:Gbu liuk am! loy,'ng i m-Oi11, S!,«1 .;;f P;rer on th~ Si\k du!
h:1 the D:Jign; tub of p«rn 1h; f'3pe, gtndy 1hcrton wilh !he H,rnd, ;uJ th<: "lrn!e Draught
will b,; imp,u\J m \C~fli.iN\l (rom th<:Gbu upor, the P,pet. . , A liiik Pu,1i£e will ttn,kr
ti"" M~lvx! o««lill8, fusil>le 3ndosy."

« The structure of representation is thus that of an infinite mise~en~abymetwith the new

apparatus having to represent the 0Id1itself representing the previous one. and so on}>,
Architecture represented by painting represented by engraving represented by book
illustration represented by photocopy represented by journal illustration. (As presen*
ted in Beaumont Newhall's History of Photography. New York, Museum of Modern
Art, 1980).
Toward a theory of the history of rcprcs.cntational technologies 115

A dialectical understanding of the relationship between technique

and technology opens up a broad range of possible connections.Just as
technology often automatizes an aceepted technique, so new techni-
ques often appear in reaction to -indeed in compensation for- the
introduction of the technologies. Consider the common thirties prac-
tice of aiming the mike at the floor, adopted to recover an omnidirectio-
nality lost with the advent of directional microphones. A similar
compensatory attempt leads directors in the fifties to break down the
wide Cinemascope screen into two or three sections, each of which
permits composition in a field more nearly similar to that of the
familiar I: 1.33frame. The industry's use of stereo for dialogue involves
the same dialectic once again. Traditional miking techniques, when
applied to the new stereo technology, tend to confuse the viewer, for
from shot to shot the same character's voice may move about in the new
sound space created by stereo. The simple solution to this quandary,
already operative in the late fifties and adopted today by nearly all users
of the Dolby system, is to record dialogue monaurally and to run it
either through the .center speaker alone, or through the center speaker
at full volume with the same signal down a few decibels through the
appropriate side speaker. Once again, history appears through the
dialectic of technique and technology, a dialectic which can be fully
appreciated only by the historian who maintains the distinction. This is
of course not to say that the distinction is always easy to make in
practice {e.g.the mid-nineteenth-century use of a head brace to secure
sharp focus combines a new technology -the brace itself- and a new
technique- the decision to use the brace for some subjects and not for
others). Nor does the technique/technology couple have any privileged
status which might give it priority over such pairings as technique/sub-
ject matter or technology /economics. The important thing to remenber
is that a dialectical understanding of history 1sdestroyed from the start
by any theory which reduces to one those practices that interact as two.
2. Basic apparatus/ Historical apparatus
In many ways, descriptive cinema theory of the seventies takes up
where prescriptive theory left off in 1930.Concerned to preserve cine-
matic purity from the threat of «theatrical» sound, Clair, Eisenstein,
Arnheim and others reflected for a limited but intense period on the
question of cinematic specificity. The seventies revived this concern in
the form of speculation on the ideological and/or psychoanalytical
ramifications of cinema's basic apparatus. The influential work of
Jean-Louis Comolli thus appears in the broader context of work by
Pleynet, Baudry, and Metz on the topic ofcinematic specificity. Widely
admired and imitated, this strain of criticism has never to my know-
116 Wck Allman

ledge been subjected to any of the fundamental criticisms to which its

basic strategy opens it/ 5.
While repeatedly criticizing Bazin for his unilinear realist view of
cinema history, Comolli nervertheless regularly reiterates two telltale
phrases which reveal his own conception of cinema history as a unified,
straight-line affair. From the very beginning of his long study, in an
introductory section entitled «The ideological place of the 'basic appa-
ratus'», Comolli borrows from Marcelin Pleynet the notion that
cinema «inherits» (I, 6) the code of Renaissance perspective, a notion
which implies the fundamental identity of the perspective practiced in
the Italian fifteeth century with that which characterizes the standard
cinema camera (the question of sound perspective never being brought
into play). Indeed, this assumption is perfectly consonant with Comol-
li's insistence on «the patient accumulation of technical processes»
whereby cinema has carried out the ideology's bidding (V, 98). Again,
the notion of «patient accumulation» implies a lack of contradiction
among all the processes so accumulated. Overall, adoption of this
additive approach to cinema history brings Comolli back surprisingly
close to the model for which he chides Bazin. Because its « basic
apparatus» doesn't change, cinema is by definition throughout its
history fundamentally self-identical.
Curiously, in another part of his study Comolli himself provides the
counter-argument to this proposition .. The third installment devotes a
long passage to the questionable practice of seeking out historical
«firsts» (close-up, pan, iris, etc.). Criticizing Mitry's search for the first
close-up, Comolli states : « No necessary equivalence links the close-up
of 1913 to those of 1960 because the relevant element of opposition is
not the parameter of enlargment in shots, but the network of diffe-
rences between the forces which determine two different moments of
film practice. These differences specifically preclude constituting
«close-up» (or traveling shots, etc.) into an historical chain and setting
them all on the same level. By founding the close-up in this way Mi try
effaces the scene of contrr1dictions where the conditions of cinemafo·
graphic significance are played out and erects instead an autonomous
series of technical processes : these teclJniques, once «invented», syste-
matized and enthroned by some pioneer (whose practice for this very
reason is not necessarily connected to that of laterfilmmakers)Jorever
remain what they were on first appearance, available once andfor all,
usable universally and out of time abstract molds whose nature,
function and meaning do not change» (lll,47).
The argument is·well stated indeed. Surprinsingly Saussurean in his

/5. I have, however, benefitted greatly from reading an unpublished paper by NoBI
Carroll on «The Specificity of Media and the Arts,>.
Toward a theory Qf the history of representationaltechnologies 117

reasoning, Comolli introduces through the notion of« contradiction»

the principle of excluded meanings on which semiotic analysis is based.
Because two «close-ups» enter into differing sets of contradictions in
different periods (Le. exist in a context of different excluded meanings,
invoke different background sets), they may not properly be said to
derive from the same category of expression. But if this is the case, then
on what grounds does Comolli recognize perspective across the ages as
the «same» thing 7 Indeed, the very notion of a «basic » apparatus of
cinema itself is compromised (quite rightly, I believe) by this argument.
Two important consideration are at stake here. First, there is the
fundamental question of the historicity of cinema itself, along with its
basic apparatus. Strikingly, Comolli himself argues that our notion of
« basic apparatus » needs to be broadened from Baudry's identification
of the apparatus with the camera and projector alone, yet he never
seems to realize that this very argument compromises the possible
existence of an apparatus basic enough to be self-identical from decade
to decade. If Comolli can by argument (or if Hollywwod can be
introducing soundJ change the definition of the basic apparatus, then
just how much force can the term « basic »retain? Second, as in the case
of the technique/technology conflation, continu.ed adherence to the
notion of an a historical« basic» apparatus robs us of the opportunity
to consider the development of cinema technology dialectically.
Notions of inheritance and patient accumulation leave little room for
an understanding of history which is process-oriented and which res-
pects the semiotic notion of excluded meanings, itself an underexploi-
ted but powerful! tool in support of a dialectical approach to history.
What would a non-additive approach to technology look like? How
might we write history, all the while respecting the notion that not even
the apparatus itself is independent, of history ? In order to answer these
questions I can do no better than to offer as an example of the type of
reading fostered by the principles enunciated above a hypothetical
account of the beginnings of perspective itself. Borrowing Pleynet's
term, Comolli asserts that cinema «inherits» the code of perspective
from painting. Whereas some would see an important difference or
even a contradiction between the notion of perspective as a technique
and perspective as embedded in a technology, Como Iiisees only conti-
nuity. Faithful to an incremental logic, Comolli simply treats each new
state as adding something to the previous one: cinema adds an automa-
tic quality to painting in the same way that a new fi!mstock adds color
or an additional track adds sound. If we move back to the« beginning»,
however, to the initial development on which cinema depends, to what
can we say that perspective is added? Is perspective simply an inscruta-
ble point of origin 7 Is it added to something else?
· The master art of the early middle ages, as Henri Focillon tells us,
118 Rkk A/Jman

was architecture/6. Sculpture, painting, mosaic, and all the so-called

decorative arts gained their right to exist, their meaning, and even their
physical support from architecture. Sculpture was thus rarely free-
standing but rather an integral part of an architectural edifice, while
painting was most commonly an ancillary art, regularly applied to
three-dimensi,;mal elements of architectural design, sculpture, or orna-
ment. Yet, because it was applied to three-dimensional surfaces, much
of the «flat », monochrome painting of the Romanesque period easily,
indeed automatically, represented depth, roundness, and a broad spec-
trum of shade\;!tones. But suppose that, for whatever reason, one were
to transfer the locus of painting from the three-dimensional aspects of
architecture to the flat wall between (fresco painting) and from there to
movable panels or canvases meant to be hung on tho.se flat walls.(easel
painting), what then? From a three-dimensional art, painting would be
reduced to two. The invention of perspective would then be anything
but an addition to a previous art, it would be a reaction against the
previous art, a compensation through technique for losses sustained in
the transition from one support to another, from one technology to
another, from one apparatus to another. Take three-dimensionality
away from painting, i.e. take away from painting the support which
serves as the early middle ages' guarantee of reality, and we find
painting imitating that support, attempting by some geometric black
magic to recreate three dimensions out of two - and in so doing
constantly depicting the architectural monuments, forms, and spaces
which made up the earlier support. The excluded element, present by its
absence in the early history of wall, panel, and easel painting, is thus the
three-dimensional architectural support. By the time cinema automa-
tizes perspective half a millennium later, there is of course no longer
any such excluded element ; instead there is the significant absence of
the immobility of subjects which nineteeth-century photography had
labored so hard to obtain.
Though these are but two hypothetical moments in the history of
perspective -the invention of perspeetive and its automatization in
cinema- their differences clearly exemplify the historicity of perspec-
tive in relationship to its apparatus, i.e. the non-identity of perspective
from one historical situation to another. In addition, by placing the
apparatus within history, we provide a buffer between technological or
technical change and the ideology which Comolli would invoke to
explain that change. Instead of relating each change directly to a single
set of ideological forces (which in Comolli often appear insufficiently
various and differentiated, thus corresponding all too well to his uni-

/6. Henri Focillon, La vie desformes, Paris, Pressesuniversitairesde France. 1942.

Nouvelle edition1 Paris, Presses Universitairede France, coJI.Quadrige, 198l.
Toward a theory of the history of representationaltechnologies 119

fied notion of perspective over the ages), we are encouraged by histori·

cal analysis of the apparatus to recognize as varied that which from an
ontological standpoint appears unified. This is of course by no means
precludes an ideological explanation of technological change. It only
means that such an explanation will necessarily relate to the apparatus
as it changes in history, and not to some putative «basic» apparatus
which has no historical existence.
3. Codes of representation/ Codes of reality
However much Comolli's general statements may imply recognition
of a single basic apparatus, his long analysis of the introduction of
panchromatic film stock in the late twenties reveals. a willingness to
historify technological change in something other than a purely linear,
additive manner. By looping back to a «previous» apparatus -
photography- in establishing the historical context for the introduc-
tion of panchro, Comolli poses the otherwise absent question of the
relationship among representational technologies, thus opening the
way toward a more general theory of representation as it relates to and
is conveyed by apparatuses located in history. The remainder of this
sectio~ sketches out such a theory.
lnlt~amous essay on «The Ontology of the Photographic Image»,
Andre Bazin identifies the «mummy complex » which lies at the origin
of representatlqnal art. The king dies. In order not to lose him comple-
tely, the priests make a death mask of his face. The king goes to his
tomb, but the people retain a representation of his features. The king is
dead; long live the «king». Art was thus born, Bazin suggests, as «the
preservation of life by a representation of life »/7. As a theory of
re-present-ation, Bazin's analysis stresses the extent of which all repre-
sentation is an attempt to compensate for the loss of something which is
no longer present, from a deceased sovereign or loved one to the distant
monument of which one brings home a postcard souvenir. Stressing the
relatively uncoded iconic mode of representation, Bazin constantly
plays up the relationship between the representation and the represen-
ted. Comolli, on the other hand, devotes his attention to analysis of the
representation itself, revealing its complex coding and its methods of
binding the spectator into a particular ideology. A propos of Bazin's
death mask he might well have pointed out the conventional reasons
for choosing to preserve the face (rather than, say, the arm pit), with its
implied function as window to the soul, or the implications of casting
the mask in a particular precious metal, symbol of royalty and durabi-

/7. Andre Bazin, «Ontologie de !'Image photographique». in Qu'est-ce•quele

cinema. Paris, Editions du Cerf. 1981. pp. 9-17.
120 Rick Altman

lity. From this point of view the seemingly uncoded nature of the iconic
mode nevertheless offers up its potentially complex coding. It is this
coding that ensures a sense of reality for the spectator while at the same
time bonding that spectator to an ideological position.
On the one side, then, the real ; on the other side its representation -
less coded for Bazin, more coded for Comolli ; leaving Bazin's specta-
tor at liberty, ·invisibly binding Comolli's. To my mind Comolli's
position is a significant gain over the Bazinian stance, yet from another
point of view the two critics are surprisingly similar in their views. In
both cases the real, that which is represented, appears as a natural fact
and not as a coded construct. To be sure, Comolli discusses at length
the codes which identify a representation as successfully representing
the real, but he never shows the least concern for the codes which mark
the real as real. The panchromatic film stock example might well have
permitted him to do so, but instead of identifying the photographic
codes with which cinema aligns itself as part of the code of reality to
which cinema must conform, he treats the photographic codes as a new
component of the code of representation. As in the case of the coming
of sound, cinematic technology is seen as responding directly and
nearly automatically to some new ideological development.
Yet the real is no less coded than representation. Let us take the
extreme example, already evoked, of perspective painting. It seems that
the rise of perspective corresponds to a desire to imitate the three
dimensions of nature on a two-dimensional plane. Yet where is it said
that nature has three dimensions? And why is it that the earliest
perspective paintings are nearly without exception of religious subjects,
usually including architectural decor? From Giotto to Lorenzetti, the
pioneers of perspective always refer to the reality that is coded by their
world. Actions have reality to the extent that they are recognizable as
deriving from a limited number of accepted texts; building have reality
to the extent that they appear to possess the three dimensions which can
be traced back to the Temple in Jerusalem; people have reality to the
extent that the culture ascribes to them that reality. Thus Virtue and
Vice exist, as do Good Government and Bad Government, but indivi-
dual portraits and townscapes of Siena as Siena will have to wait a
century. Well known individuals and cities may serve as models for
disciples or principles, but only later will their success as representa-
tions depend on the personal resemblance rather than resemblance to a
well coded, previously established category (at which time the very
notion of« personal resemblance» will be subjected to a coding particu-
lar to its age).
But what difference does it make that reality should be coded? And
what is the source of that coding? The answer to the latter question is
obvious and would no doubt have occurred to Comolli had he not
120 Rick Altman
lity. From this point of view the seemingly uncoded nature of the iconic
mode nevertheless offers up its potentially complex coding. It is this
coding that ensures a sense of reality for the spectator while at the same
time bonding that spectator to an ideological position.
On the one side, then, the real ; on the other side its representation -
less coded for Bazin, more coded for Comolli ; leaving Bazin's specta­
tor at liberty, ·invisibly binding Comolli's. To my mind Comolli's
position is a significant gain over the Bazinian stance, yet from another
point of view the two critics are surprisingly similar in their views. In
both cases the real, that which is represented, appears as a natural fact
and not as a coded construct. To be sure, Comolli discusses at length
the codes which identify a representation as successfully representing
the real, but he never shows the least concern for the codes which mark
the real as real. The panchromatic film stock example might well have
permitted him to do so, but instead of identifying the photographic
codes with which cinema aligns itself as part of the code of reality to
which cinema must conform, he treats the photographic codes as a new
component of the code of representation. As in the case of the coming
of sound, cinematic technology is seen as responding directly and
nearly automatically to some new ideological development.
Yet the real is no less coded than representation. Let us take the
extreme example, already evoked, of perspective painting. It seems that
the rise of perspective corresponds to a desire to imitate the three
dimensions of nature on a two-dimensional plane. Yet where is it said
that nature has three dimensions? And why is it that the earliest
perspective paintings are nearly without exception of religious subjects,
usually including architectural decor? From Giotto to Lorenzetti, the
pioneers of perspective always refer to the reality that is coded by their
world. Actions have reality to the extent that they are recognizable as
deriving from a limited number of accepted texts; building have reality
to the extent that they appear to possess the three dimensions which can
be traced back to the Temple in Jerusalem; people have reality to the
extent that the culture ascribes to them that reality. Thus Virtue and
Vice exist, as do Good Government and Bad Government, but indivi­
dual portraits and townscapes of Siena as Siena will have to wait a
century. Well known individuals and cities may serve as models for
disciples or principles, but only later will their success as representa­
tions depend on the personal resemblance rather than resemblance to a
well coded, previously established category (at which time the very
notion of« personal resemblance» will be subjected to a coding particu­
lar to its age).
But what difference does it make that reality should be coded? And
what is the source of that coding? The answer to the latter question is
obvious and would no doubt have occurred to Comolli had he not
Toward a theory of the history of representational technologies 121
limited his attention almost entirely to a single expression of Western
ideology. Only by stressing perspective nearly exclusively is Comolli
iIi able to make the history of Western apparatuses from the Quattrocento
to the tw�ntieth ce�tury seem _to nearly s_traight:li_n e. For if each appa­
ratus defines a particular version of reality, cod1f1es the systems requi­
red for successful representation, thus establishing the necessary and
sufficient conditions for representing the real, then we must conclude
that each apparatus establishes the code of reality to which the subse­
quent apparatus must adhere. Perspective painting imitates architec­
ture and sacred narrative, for those are the privileged apparatus
-Focillon's master art- of the early middle ages. With panchromatic
film and the addition of sound, film is responding to the definition of
reality propounded by its three most immediate predecessors and early
competitors - photography, radio and the theater. In order to repre­
sent properly, each new technology must therefore succeed in represen­
ting not reality itself, but the version of reality established by a
previously dominant representational technology.
In other words; there is no such thing as representation of the real ;
there is only representation of representation. For anything that we
would represent is already constructed as a representation. The struc­
ture of representation is thus that of an infinite mise-en-abyme, with the
new apparatus having to represent the old, itself representing the
previous one, and so on. Each new apparatus might thus be likened to a
translation. Expressive of an ideology different from that of the former
apparatus, the ·new apparatus must simultaneously find a way to
express that new ideology and -in the same words, as it were-seem to
be expressing the old. For only with this appearance of translation can
the new apparatus be taken as representing the real. The new system
must thus -at least provisionally- speak with two voices or risk
failure. It must sound like the old, and yet be new. When a change of
apparatus appears to involve nothing more than an addition to the old,
as in the case of sound film, the operation is carried out with little help
from the techniques deployed by the texts for which the new apparatus
serves as vehicle. When the change of apparatus is as radical as the
move from architectural three-dimensional painting to a flat surface,
then only by extraordinary technical developments, like that of pers­
pective, can the new apparatus retain its right to representation.
Now the process of translation, as everyone knows, can never pre­
iend to complet�n�ss. From _the very fact that each language is an
independent sem10t1c system, 1t would appear that every translation is
by definition partial. Part of the message may be retained, but part will
also be left behind. This essential characteristic is perfectly visible in the
development of photography. In order for a photograph to serve as an
adequate representation in nineteeth-century Europe, it had to satisfy
122 Rick Altman

the codes of the real developed by painting and drawing; that is it

required a strong iconic resemblance, a sense of depth, and a treatment
of light consonant with that depth, as well as a certain palette of colors.
But a photograph also had to satisfy the codes of representation
imposed by an industrialized bourgeoisie; that is it must require mini-
mal human talent, it must be produced mechanically, and it must be of
such a size and durability to encourage rapid and continuous sales to a
large group of modest customers (as opposed to the small group of rich
customers associated with painting). Photography succeeded because,
with one exception, it satisfied the (previous) codes of reality as well as
the new codes of representation associated with a current ideology. As
effort went into satisfying the one remaining requirement -color-
further shortcomings were noted, such as lack of durability as compa-
red to painting. But this feature, which appears as a drawback in the
reality code, turns out to be an advantage in the throwaway market of
the representational code. Graininess, blurring, lack of depth of focus ·
-all are slowly «corrected» as the century goes by, thus increasingly
aligning photography with the reality codes of painting.
The transition from painting to photography might thus be likened
to an imperfect translation-in-progress. Unable to express exactly the
reality codes of the preceding medium, photography is characterized by
a constant effort to make up for this lack. For the lack of color isfelt as
a lack, and not in comparison to nature, but in comparison to painting
(as well as tinted engraving and other related technologies). As long as
the new medium appears to society as in some sense competing with the
old, then the imperfection of the translation becomes a sign within the
translation itself. In any semiotic system the meaning of a given term is
defined by its actual syntagmatic relations and its potential paradigma-
tic relations. Now, it might seem that the colorless character of early
photography remains uncommuted as a textual sign by virtue of the
absence of a color paradigm in photography, but this would be true
only in the limits of photography in fact the limits of the signifying
system in question. In fact, I would claim, the appearance of a new
apparatus and textual system always introduces a period, which might
last for days or decades, when the semiotics of the new system must be
seen as including those of the old system, which it translates. In short,
the lack in the new system -say color in early photography- must be
read as a sign in that system even though it appears only by its absence.
Looked at in another way, color might be considered as a signified
inherited from the previous system of representation (now serving as
the new code of reality), but for which the present system provides no
signifier. This imbalance in the system creates a pressure of two kinds;
I) to adapt the apparatus so that it can signify in the desired way, and 2)
Toward a theory of the history of representational technologies 123

to adapt the texts deployed by the new apparatus to the desired

signification.
When cinema was invented, for example, it had at least the photogra-
phic tradition to contend with, itself the representationofarepresenta-
tion and thus encapsulating the remnants of prior systems as well. To
reduce the problem to its very skeleton, we might say that cinema
produced a supplement of movement while relinquishing the possessi-
bility of the representation -a trait which photography intensifies as
compared to the rarer ownership of paintings, which were in turn more
commonly owned by individuals than the early medieval art work.
Now cinema adapts to this lack in two ways. First, unable to offer
possession of the apparatus to the consumer, cinema offers tales of
possession, the dominant plot concern of early cinema. Second, the
cinema industry spawns a secondary market of possessible artefacts:
fan magazines, pictures of stars, autographs, souvenir programs. Only
with the video medium will it finally become convenient and common
actually to own one's feature film.
Indeed, the full-scale quotation of one medium by another (exempli-
fied by the feature film video cassette) leads me to a final point. The
easiest way to prove that a new medium meets the reality demands
defined by the old is simply to quote the old, verbatim, as it were.
Perspective painting quotes architecture, printing quotes speech, early
printed stories portray oral story-telling, Renaissance theater uses a
perspective set, early film records theatrical performances, and so on.
As a special and particularly obvious case ofrepresentation's perpetual
status as the representation of a representation, these examples remind
us of the extent to which Octave Mannoni's celebrated description of
the dynamics of spectatorship apply as well to the dynamics of the
history of representation. «Je sais bien, says Mannoni, I know that this
text is only a representation;« mais quand-meme », but still, I have my
own reasons for believing. The same mechanism operates in the play of
representational systems. I know that this new medium is not the same
as the old, but it flatters my needs, it does what I need it to do. In short,
the reality codes may be slightly off, but the ideology -the representa-
tional code- is right.

Consider RCA's recent videodisc commercials. We watch a family

choose which film they want to watch, implying that the magic video-
disc box can quote the film verbatim. Now we know that a reduced-size
film is not a real film (it's cropped, it has low fidelity, and so forth). But
we also know that we -as the TV family members are quick to point
out- are watching the commercial while they are watching their
chosen film. And so it happens that we come to desire an RCA
videodisc player. This constant process of quotation and erasure, as
124 Rick A ltmun

Jane Feuer has called it/8, serves alternately to point out the similarities
between the two media and then to erase them, in the process aligning
the new representational mode on the old codes ofreality, then offering
an ideological plus which forever sets the new mode over the old. I
know it's not a pristine 35mm print, but what the key, there are no
commercials and I can stop for a beer whenever I want.
The straight-line model assumed throughout this paper is of course
used here only for the sake of presentational convenience. There is no
single straight line from the Ark of the Covenant passing through
Assisi, Quattrocento perspective painting, Renaissance and neo-
classical theater, photography, cinema, and TI. Instead, there is a
complex web of constantly changing relationships.among representa-
tional technologies. The challenge of the history of representation, as I
have sketched is here, is in the task of identifying the return of one
repressed representational system in another, and thus in observing the
unceasing pressure to which media subject each other. Seen as a system
with its own history and internal dynamics, the ideology of representa-
tion opens itself up to the kind of historical analysis in which the system
itself, once launched, must be seen as retaining a certain life of its own.
Changes in external ideology occur as pressures on the system, espe-
cially in the form of new codes of representation, but external ideology
is no longer the only thing driving the system. New ideologies cannot
simply generate new representational systems without taking into
account the reality codes established by previous and/ or competing
representational systems. Once again, we are led to a type of history
which remains fundamentally dialectical. Instead of seeing a straight
line between an initial ideological impulse and an ultimate technologi-
cal development (as does Comolli), this new approach considers that
every ideological force must by necessity grapple with the residue of
another ideological impetus embodied in competing representational
modes. To write the history of representational technologies is thus to
trace the dialectic which grows out of the confrontation between repre-
sentational and reality codes.

/8. Jane Feuer, The Hollywood Musical, Bloomington, Indiana Univ. Press, 1982.
124 Rick Altman Toward a theory of the history of representational technologies 125

Jane Feuer has called it/8, serves alternately to point out the similarities Revenant sur /es rapports technique/ideologie don/ /'analyse avail ete
between the two media and then to erase them, in the process aligning initiee en par tie par J. L. Comolli a partir des travaux sur l'apparei/ de
the new representational mode on the old codes of reality, then offering base de Jean-Louis Baudry et Marcel/in Pleynet, ce texte vise a re­
an ideological plus which forever sets the new mode over the old. I historiser et a dialectiser ces conceptions qui re/event d'une histoire par
know it's not a pristine 35mm print, but what the key, there are no trop lineaire et accumulative. Revenant sur /'opposition, la plupart du
commercials and I can stop for a beer whenever I want. temps neutra/isee entre technique et technologie, ii introduit la notion
d'appareil historique. II etudie la nature des liens non seulement entre
The straight-line model assumed throughout this paper is of course /es codes de representation et /'appareil de base mais aussi entre celui-ci,
used here only for the sake of presentational convenience. There is no /es codes de representation et /es codes historiques qui definissent
single straight line from the Ark of the Covenant passing through /'apprehension de la dite rea/ite.
Assisi, Quattrocento perspective painting, �enaissance and neo­
classical theater, photography, cinema, and TV. Instead, there is a
complex web of constantly changing relationships.among representa­
tional technologies. The challenge of the history of representation, as I
have sketched is here, is in the task of identifying the return of one
repressed representational system in another, and thus in observing the
unceasing pressure to which media subject each other. Seen as a system
with its own history and internal dynamics, the ideology of representa­
tion opens itself up to the kind of historical analysis in which the system
itself, once launched, must be seen as retaining a certain life of its own.
Changes in external ideology occur as pressures on the system, espe­
cially in the form of new codes of representation, but external ideology
is no longer the only thing driving the system. New ideologies cannot
simply generate new representational systems without taking into
account the reality codes established by previous and/ or competing
representational systems. Once again, we are led to a type of history
which remains fundamentally dialectical. Instead of seeing a straight
line between an initial ideological impulse and an ultimate technologi­
cal development (as does Comolli), this new approach considers that
every ideological force must by necessity grapple with the residue of
another ideological impetus embodied in competing representational
modes. To write the history ofrepresentational technologies is thus to
trace the dialectic which grows out of the confrontation between repre­
sentational and reality codes.

/8. Jane Feuer, The Hollywood Musical, Bloomington, Indiana Univ. Press, 1982.